Investigation report Report Report title Activity number

Undesirable incident with HTV Eagle pipehandling crane – Statoil

Gullfaks B – 7 March 2017

001050062

Security grading

Public

Not publicly available

Restricted

Confidential

Strictly confidential

Summary

The incident occurred at 15.25 on 7 March 2017 on the Gullfaks B (GFB) installation. The boom on

the horizontal to vertical (HTV)) Eagle pipehandling crane, which weighed about 14.4 tonnes, fell 10

metres to the pipe deck after the steel rope failed. Two people had adjusted the grippers on the crane’s

gripper yoke down on the pipe deck immediately before the incident occurred.

Drilling facilities on GFB were upgraded with a new HTV Eagle crane in 2015. This was part of a big

project executed for Statoil by Aibel. MH Wirth was Aibel’s subcontractor for the pipehandling

equipment. Operations resumed after the upgrading in September 2015. KCA Deutag (KCAD) is the

drilling contractor on GFB.

The direct cause of the crane boom falling down was fatigue in the steel rope. This was one of the

main components in the arrangement for the crane’s vertical up and down movement.

The underlying cause was weaknesses in the design of the crane’s hoisting arrangement, which

consists primarily of the winch, sheaves, vertical dolly and steel rope. Owing to the design, the rope

was subject to wear and fatigue over time. Fatigue in the steel rope was not assessed as a relevant

problem for this design.

Risk associated with the equipment was not adequately identified, replacement intervals were not

determined on the basis of durability calculations for the steel rope, and insufficient use was made of

crane and lifting expertise when specifying the crane requirements. In addition, follow-up of earlier

incidents involving damage to steel rope and a failed sheave bearing was inadequate, deficiencies and

errors existed in the maintenance programme for the rope, and follow-up and checking when taking

delivery of equipment and machinery were inadequate.

Under slightly different circumstances, the incident could have caused serious personal injuries or

loss of human life.

The PSA team’s job was to support the police inquiry and to conduct its own investigation of the

incident. The PSA and the police were on board GFB from 8–10 March 2017.

Involved Main group Approved date

T-1 24 August 2017

Members of the investigation team Investigation leader

Anne Marit Lie, Reidar Sune, Gustav W Dunsæd,

Eigil Sørensen, Sigmund Andreassen (did not

participate offshore)

Anne Marit Lie

2

Contents Summary .................................................................................................................................... 3 1 Introduction .......................................................................................................................... 3

Mandate for the investigation ........................................................................ 4 Composition of the investigation team .......................................................... 4

2 Background information ...................................................................................................... 4 3 Course of events ................................................................................................................... 7 4 Direct and underlying causes of the incident ..................................................................... 12

4.1 Direct cause ......................................................................................... 12 4.2.1 Design of the HTV Eagle crane ............................................. 12 4.2.2 Failure to follow up earlier incidents ..................................... 15 4.2.3 Insufficient knowledge of and expertise on steel rope ........... 16 4.2.4 Monitoring/inspection of steel rope ....................................... 16

5 Potential of the incident ...................................................................................................... 17 5.1 Actual consequence ............................................................................ 17 5.2 Potential consequence ......................................................................... 18

6 Observations ....................................................................................................................... 18 6.1 Nonconformities .................................................................................................... 19

6.1.1 Risk assessment of equipment ................................................ 19 6.1.2 Investigation of and improvement measures following

earlier incidents ..................................................................... 19 6.1.3 Utilisation of expertise ........................................................... 20 6.1.4 Responsibility for taking delivery and operation of

equipment ............................................................................... 21 6.1.5 Maintenance ........................................................................... 22

7 Poor visibility from crane cabin ...................................................................... 22 8 Barrier assessment .............................................................................................................. 22 9 Discussion of uncertainties ................................................................................................. 23 10 Statoil’s investigation report ............................................................................................ 24 11 Appendix .......................................................................................................................... 24

3

Summary

The incident occurred at 15.25 on 7 March 2017 on the Gullfaks B (GFB) installation. The

boom on the horizontal to vertical (HTV)) Eagle pipehandling crane, which weighed about

14.4 tonnes, fell 10 metres to the pipe deck after the steel rope failed. Two people had

adjusted the grippers on the crane’s gripper yoke down on the pipe deck immediately before

the incident occurred.

Drilling facilities on GFB were upgraded with a new HTV Eagle crane in 2015. This was part

of a big project executed for Statoil by Aibel. MH Wirth was Aibel’s subcontractor for the

pipehandling equipment. Operations resumed after the upgrading in September 2015. KCA

Deutag (KCAD) is the drilling contractor on GFB.

The direct cause of the crane boom falling down was fatigue in the steel rope. This was one of

the main components in the arrangement for the crane’s vertical up and down movement.

The underlying cause was weaknesses in the design of the crane’s hoisting arrangement,

which consists primarily of the winch, sheaves, vertical dolly and steel rope. Owing to the

design, the rope was subject to wear and fatigue over time. Fatigue in the steel rope was not

assessed as a relevant problem for this design.

Risk associated with the equipment was not adequately identified, replacement intervals were

not determined on the basis of durability calculations for the steel rope, and insufficient use

was made of crane and lifting expertise when specifying the crane requirements. In addition,

follow-up of earlier incidents involving damage to steel rope and a failed sheave bearing was

inadequate, deficiencies and errors existed in the maintenance programme for the rope, and

follow-up and checking when taking delivery of equipment and machinery were inadequate.

Under slightly different circumstances, the incident could have caused serious personal

injuries or loss of human life.

The PSA team’s job was to support the police inquiry and to conduct its own investigation of

the incident. The PSA and the police were on board GFB from 8–10 March 2017.

Statoil conducted its own investigation.

1 Introduction

In connection with transferring materials from the pipe deck to the drill floor on GFB, the

crane boom dropped 10 metres without warning to the pipe deck. Statoil notified the PSA of

the incident at 17.55 on 7 March 2017.

The police requested support from the PSA for its inquiry into the incident. At the same time,

the PSA conducted its own investigation.

Four members of the PSA’s investigation team were on board GFB together with the police

from 8–10 March 2017. During this time, the team supported the police during site

inspections and in interviews with personnel.

Following its stay on GFB, the team received further information from interviews with

relevant personnel at KCAD, MH Wirth, subcontractors involved and Statoil.

4

The steel rope and one sheave were confiscated by the police. By agreement with the police

and the PSA, Statoil sent the confiscated components to W Giertsen Services AS in Bergen

for assessments and analyses.

Statoil’s investigation report was received by the PSA on 8 June 2017. A meeting was held

with Statoil on 19 June 2017 to clarify certain points in this document.

The investigation team has utilised the human, technological and organisational (HTO)

approach as its methodology. This builds on the concept of operational, organisational and

technical barrier elements.

Mandate for the investigation

a. Clarify the incident’s scope and course of events with an emphasis on safety, working

environment and emergency preparedness aspects.

b. Assess the actual and potential consequences

1. Harm caused to people, material assets and the environment.

2. The potential of the incident to harm people, material assets and the

environment.

c. Assess direct and underlying causes, with an emphasis on human, technology and

organisation (HTO) and operational aspects, from a barrier perspective.

d. Discuss and describe possible uncertainties/unclear aspects.

e. Identify nonconformities and improvement points related to the regulations (and

internal requirements).

f. Discuss barriers which have functioned (in other words, those which have helped to

prevent a hazard from developing into an accident, or which have reduced the

consequences of an accident).

g. Assess the player’s own investigation report (the PSA’s assessment is to be

communicated in a meeting or by letter).

h. Prepare a report and a covering letter (possibly with proposals for the use of

reactions) in accordance with the template.

i. Recommend – and contribute to – further follow-up.

Composition of the investigation team

The PSA’s investigation team has comprised:

Anne Marit Lie – investigation leader, logistics and emergency preparedness

discipline area

Reidar Sune, logistics and emergency preparedness discipline area

Gustav W Dunsæd, drilling and well discipline area

Eigil Sørensen, drilling and well discipline area

Sigmund Andreassen (did not participate offshore), logistics and emergency

preparedness discipline area

2 Background information

5

Gullfaks is a Statoil-operated field located in 130-220 metres of water in the northern North

Sea. It has been developed with three installations, each featuring integrated process, drilling

and quarters modules, a concrete support structure and a steel topside. Oil and gas from GFB

are transferred to Gullfaks A and C for processing and storage. The oil is loaded into shuttle

tankers on the field, while the gas gets piped to the processing facilities at Kårstø.

GFB is a quarters, drilling and oil production/water injection facility of the Condeep type,

standing in 145 metres of water.

Activities before the incident

- Well B-14C was drilled to a depth of 5 145 metres.

- Seven-inch casing was installed at 5 144 metres and cemented.

- The well was tested to 237 bar using mud with a specific gravity of 1.65.

- The well and blowout preventer were tested (no drilling had taken place out of the

seven-inch liner).

Pipehandling equipment involved in the incident

The pipehandling system primarily comprises a traverse crane on the pipe deck and an

HTV Eagle crane installed in connection with the V-door to the drill floor. Drill pipe is

transported horizontally by the traverse crane from the pipe deck to the catwalk for further

handling by the HTV Eagle crane to and from the drill floor.

The HTV Eagle crane comprises a vertical dolly, which is installed between two vertical

guide beams for up and down movement with the aid of a hoisting arrangement, and equipped

with a boom. Its role is to hoist drill pipe from the catwalk on the pipe deck, where the pipes

lie horizontally, to a vertical position on the drill floor.

In the hoisting arrangement, steel rope from an electric winch runs over a sheave at the top of

one guide beam and under two sheaves on the dolly to its anchoring point on top of the other

beam. A load cell is installed as part of the anchoring arrangement. See figures 1 and 2 for the

arrangement of the HTV Eagle crane.

6

Figure 1 HTV Eagle arrangement, with the steel rope in red. (Statoil’s investigation report)

Figure 2 The HTV Eagle crane in the derrick and the traverse crane on the pipe deck.

HTV Eagle crane

Catwalk

Traverse crane

7

3 Course of events

A new HTV Eagle crane was installed in 2015 as part of a major conversion of the drilling

facilities on GFB. Statoil outsourced the work to Aibel in 2013, which awarded the

subcontract for the HTV pipehandling equipment (specified as the HTV Eagle crane and

tubular feeding machine) to MH Wirth. The latter bought the winch with drum from

Kverneland Aqua AS.

When designing the HTV Eagle crane, MH Wirth carried out a criticality assessment using

failure mode, effects and criticality analysis (FMECA). But this analysis failed to identify the

steel rope as a critical component in the system.

The GFB project drew on experience with the HTV Eagle crane on Gullfaks C in an effort to

improve the design. Emphasis was given, for example, to protecting the steel rope in the

hoisting arrangement from being damaged by the deck crane when the latter was used to lift

loads in and out through the V-door to the drill floor. On Gullfaks C, the rope had previously

been damaged during material handling by the offshore crane.

To improve the design, the GFB project team sought to give the rope more protection by

incorporating it in the guide beams. That was also why a sheave diameter 18 times greater

than the rope diameter (18xD/d) came to be chosen, since this dimension was better adapted

to the design while also simplifying the structure.

Where safety aspects of the HTV Eagle design are concern, the primary emphasis appears to

have been on preventing loads and objects dropping from the crane and on ensuring that drill

pipe could not slip out of the gripper yoke.

Following a review of the documentation and conversations with personnel involved at

Statoil, MH Wirth and KCAD, the team’s understanding is that no particular attention was

devoted to the hoisting arrangement with regard either to the choice of rotation-resistant steel

rope or the assessment of expected service life for this rope in the relevant winch and sheave

arrangement. The choice of rotation-resistant rope was not based on assessments of several

rope types.

MH Wirth issued a conformity declaration for the HTV Eagle crane on 16 July 2015.

DNV GL issued a certificate for the HTV Eagle lifting facility/equipment on 5 September

2015. This was based on a low-level verification as defined in DNV-OSS-308, which is used

for lifting facilities/equipment with a low risk level. The certification identified no design

weaknesses or conditions to be kept under observation in the operating phase.

The investigation identified inadequate spare-part lists and maintenance routines at the point

of delivery from project to operations. No documentation has been found which shows that

maintenance and inspection routines have taken account of the fact that rotation-resistant steel

rope is difficult to inspect visually for fatigue. Such conditions should have been covered in

the manuals from MH Wirth.

The crane became operational in September 2015. Before start-up, Statoil gave emphasis to

educating and training the personnel who were to use new equipment.

8

Incident January 2016: damage to steel rope

Personnel discovered damage to the steel rope on the HTV Eagle crane on 1 January 2016

(Synergi # 1461090).

Four of 34 strands proved to have parted, and the rope was deformed. A birdcage had formed.

See figure 3. The reason for this damage was assessed to be the wrong type of rotation-

resistant rope. While rope with a right-hand lay was used on the winch drum, this was

designed to handle left-hand lay.

MH Wirth drafted a major operational failure report on 10 January. The rope was replaced

with a left-hand lay type, and the problem was regarded as solved. MH Wirth ordered new

rope from Kverneland Aqua AS, which received this from Westfalische Drahtindustrie.

MH Wirth asked Kverneland Aqua AS to assess the possible cause of the damage to the rope.

The latter did as requested and submitted its report to MH Wirth. Dated 12 January, this

recommends that the hoisting arrangement be designed with a sheave diameter of 26xD/d and

warns that the service life of the steel rope would be shortened by using sheaves with small

dimensions.

Figure 3 Damage to the steel rope on the HTV Eagle crane. (Synergi # 1461090, January 2016)

In an asset integrity report of 17 January, the KCAD field engineer for crane and lifting expresses concern that the fleet angle of the steel rope from the electric winch drum to the upper sheave (S1, see figure 4) atop the one guide beam is at the upper limit of the tolerable and must be followed up. MH Wirth responded that the fleet angle was within the sheave’s tolerance.

On 19 January, MH Wirth submitted a technical report on the damage to the rope, where it maintains the view that this was caused by using rope with right-hand instead of left-hand lay. The report did not incorporate the assessments provided by Kverneland Aqua AS in its report of 12 January.

MH Wirth made no reassessment after receiving the input from Kverneland Aqua AS and KCAD. No service life was calculated for the rope on this occasion either.

9

Incident of July 2016

On 30 July, personnel heard strange noises from the HTV Eagle crane (Synergi # 1480757).

These proved to originate not from the winch but from the S1 sheave at the top of the guide beam. Soon afterwards, while the crane was in use, rollers from the sheave bearing fell to the winch deck.

ß

Figure 2 The HTV Eagle crane when bearings were replaced in S1. (Synergi # 1480757)s

The bearing of S1 on the guide beam was replaced on that same day.

KCAD submitted a major operational failure report on 1 August covering the damage to the

bearing in S1. This identified the fleet angle of the steel rope from the drum to the sheave as

the cause of the roller bearing failure. Furthermore, the sheave was to be replaced every six

months (established as the maintenance interval).

Upper sheave (S1), where rollers fell from the

bearing

Position of the winch for hoisting the crane

Two sheaves on the dolly (S2 and S3)

10

Figure 5: The steel rope bearing Figure 6: The fleet angle between drum and sheave.

against the sheave.

Statoil’s asset integrity unit and the crane and lifting supervisor in the process section of GFB

expressed concern in an e-mail of 19 October over the design of the HTV Eagle crane. They

stated in part: “Several conditions exist here, [such as] bearing failures and abnormal sheave

wear, which derive from a large fleet angle. If nothing is done, failure of sheave bearings or

the actual sheave, or a new incident with the steel rope, can be predicted with certainty.”

This expression of concern went to Statoil’s rig representative and the KCAD maintenance

manager. It had not led to any special follow-up by either Statoil or KCAD before the rope

failed.

Measures taken

A new sheave was installed on the HTV Eagle crane in January 2017, with a visual inspection

carried out for the rope. The latter was re-installed after the inspection because no external

damage was found. Possible fatigue damage to the rope was not assessed.

Course of events, 7 March 2017

15.05. The HTV Eagle crane is used to lay steel rope offcuts on the deck after a slip and cut of

drill line operation.

15.10–15.11. Two men stood under the crane and disconnected the load on the deck.

15.12–15.13. A new load was connected – a magnet mounted on a short length of 3 ½-inch

drill pipe.

15.14–15.20. The load was readied, and the magnet lifted to the drill floor.

15.21. The magnet was unhooked on the drill floor.

15.22. The crane boom was lowered to the deck without a load.

15.23. Two people stood under the crane and adjusted the grippers to 3 ½-inch drill pipe.

11

15.24. The gripper yoke was readied, and the crane hoisted up to give the traverse crane

access to the pile of drill pipe on the deck.

15.25. Without warning, the steel rope failed and the crane boom fell 10 metres to the

pipe deck.

15.27–15.28. The borehole was covered and the platform management notified.

15.30–15.40. A site inspection was conducted, the area cordoned off was expanded, all work

permits were withdrawn and all work was halted.

15.40. All personnel involved from the drilling contractor were assembled in the coffee bar.

The emergency response organisation was not mobilised.

Figure 7 presents the approximate position of the steel rope when it failed.

Figure 7 The red arrow shows the point where the steel rope failed.

Upper sheave (S1), where rollers fell from

the bearing

Two sheaves (S2 and S3) on the dolly

Failure site

12

4 Direct and underlying causes of the incident

4.1 Direct cause

The direct cause of the boom falling to the pipe deck was fatigue in the steel rope. The rope is

a safety-critical main component in the arrangement for the up and down crane movement. It

carries the whole weight of the crane’s hoisting and boom arrangement – a maximum of 17.9

tonnes with load and about 14.4 tonnes without load.

The load on the hoisting and boom arrangement can be reduced by parking in a cradle on the

pipe deck, but use of the parking cradle depends on the position of the derrick. When the

incident occurred, the cradle was not accessible because of the derrick’s position in relation to

the pipe deck. In order to use the traverse crane on the pipe deck, the HTV Eagle crane boom

had to be parked in a position above the V-door and drill floor.

When the HTV Eagle crane is parked above the pipe deck, the boom is held in place only by

the rope. The latter will then be subject to a continuous load of about 14.4 tonnes. In addition

comes a good deal of motion and vibration in the derrick during drilling, which could be

unfavourable for the rope in terms of fatigue. That has not been included as a contributory

factor in the investigation.

The crane was not provided with any form of safety system which prevented its hoisting and

boom arrangement from falling if the steel rope or other load-bearing components failed. Nor

had any arrangement been installed which hangs off or relieves the load on the rope when the

crane is parked in a position above the pipe deck. A parking cradle is used when the crane is

parked on a level with the pipe deck.

4.2 Underlying causes

4.2.1 Design of the HTV Eagle crane

The PSA’s view is that fatigue occurred in the steel rope because of the inadequate design of

the hoisting arrangement for up and down movement of the crane. During the design phase,

assessment errors were made in choosing component dimensions and insufficient attention

was paid to how these would function in combination with the steel rope. This has result in a

very low service life for the rope. No service life calculations were made for it.

This applies to the following components.

Rotation-resistant steel rope

According to the technical literature and supplier information, using rotation-resistant steel

rope in this type of hoisting arrangement has little point in design terms. Such rope has

primarily been developed for systems with a free end, where it is desirable to reduce rotation

in the latter. Rotation-resistant steel rope is commonly used in a crane or hoisting winch

where the smallest possible rotation of the crane hook is desirable. However, this was not a

problem in the GFB system since the rope is permanently anchored at both ends (in the winch

drum and at the end attachment). Providing the rope is correctly installed in such systems, no

further rotation of the rope will be introduced and rotation will not be a challenge. No design

reason exists for choosing rotation-resistant steel rope in this type of hoisting arrangement.

13

The steel rope in use when the incident occurred had a tensile strength of 2 160 N/mm2. That

gives it high strength, but is unnecessary to meet the rope’s required safety factor. Steel rope

with such a high tensile strength is less elastic and unsuitable for a sheave with a diameter as

small as 18xD/d, which will reduce its fatigue life even further. Product information from the

manufacturer of the relevant rope recommends a minimum sheave diameter of 20xD/d.

According to the report from W Giertsen Service AS, the theoretical applied working life

would be about 50 per cent of the calculated service life (until fatigue/strand fracturing occurs

internally). Internal strand fracturing means the rope must be scrapped, see ISO 4309, while

scattered external strand fractures can be tolerated. See the table in ISO 4309. Only three

external scattered strand fractures within a length of 6xd are permitted for this type of steel

rope. The maximum applied service life for this rope is therefore theoretically about seven

months, assuming that the crane has been in normal use for 15 months,

The investigation found that MH Wirth had not documented requirements for steel rope and

the type of rope the system was designed for. No service life calculations were carried out for

the system. Nor was this issue covered in the user or maintenance manuals, or in other project

documents.

Sheave with a diameter of 18xD/d

It emerged in interviews that experience from Gullfaks C called for the steel rope to be

protected inside the guide beams to avoid being damaged during material handling when the

deck crane was used for lifts to the pipe deck. As a result, 18xD/d was chosen because it fitted

well with the design and dimensioning of the beams. This choice also accords with the

recommendation in DNV GL 2.22 that 18xD/d can be used as the smallest sheave diameter.

The standard is normative and fairly general and has failed, among other considerations, to

address complicated assemblies and arrangements involving the use of multiple sheaves and

where the steel rope could be led over several sheaves, change direction several times and so

forth. Special assessments must be made where such designs are concerned. The standard was

probably not entirely suitable for this particular design.

Winch with a drum diameter of 18xD/d

It emerged during interviews with MH Wirth that the choice of drum diameter was related to

the sheave diameter. The use of a drum diameter of 18xD/d is unlikely to have made any

appreciable contribution to steel rope fatigue, since the latter runs in and out of the drum less

frequently than over the sheaves.

Positioning of the winch

The winch drum is positioned at right angles to the sheaves. This reflects limited space and

the need to avoid blocking an exit from the drilling module. It has resulted in an unfortunate

fleet angle between winch drum and S1, which has contributed to fatigue in the rope and the

load on the sheave. The fleet angle has a bigger impact on rotation-resistant rope than on

traditional types. It was 1.9 degrees for S1, and the winch drum was offset in relation to this

sheave. In addition, a fleet angle was introduced by the positioning of the winch drum and the

rope track within it, so that the total fleet angle was considerably larger than 1.9 degrees. This

meant that the point on the rope subject to the biggest load (where the break occurred) has

been subject to a fleet angle considerably larger than 1.5 degrees, which is the maximum

angle specified by the manufacturer for the relevant rope.

14

Winch/sheave arrangement and load on the steel rope

The average service life for steel rope will depend on the number of sheaves and the spooling

direction in a hoisting arrangement. See figure 8.

Figure 8 An example of the average service life of steel rope, depending on the number of sheaves and the

direction of the rope. (Casar)

The winch drum on the HTV Eagle was underspooled in relation to S1, which was

underspooled and thereby gave rise to an S bend (reversed bend) on this sheave. S2 was also

underspooled, and the rope acquired a further S bend from S1. S3 is spooled the same way as

S2, producing a total of two successive S bends for S2 and S3. See figure 9.

Figure 9 The winch arrangement for the HTV Eagle crane. (Statoil investigation report)

Key: Endefeste = end attachment; Last = load

According to W Giertsen Service AS, it is unfortunate that the angle between S1 and the drum

means that the rope is only in contact with one side of S1. That creates rotation in the rope,

which reduces service life while greatly enhancing the threat of a kink (birdcage/twist),

particularly with slack rope. While the winch is underspooled and S1 is overspooled, the big

15

distance involved (25 metres) probably means that this is not an issue. However, a much

greater problem is the S bend between S1 and S2, where the fatigue/break occurred. Steel

rope which lifts its maximum load (14.4 + 3.5 tonnes) continuously is more exposed to

reduced service life in the event of an S bend than on equipment where the load varies from

zero to the maximum safe working load (SWL). See the report from W Giertsen Service AS.

The combination of rotation-resistant steel rope of high tensile strength, led over multiple

sheaves of a small diameter (18xD/d) and subject several times to S bends during each crane

operation, with a relatively high fleet angle between winch drum and upper sheave has

resulted in loads which have resulted in a short service life for the rope. No service life

calculations were carried out for it. The investigation found that the fatigue occurred in the

area subject to the largest number of passes over several sheaves during each work operation

by the crane.

Conversation with MH Wirth revealed that components in the hoisting arrangement were

largely chosen on the basis of the DNV GL 2.22 standard for certification of lifting appliances

and of products delivered earlier which involved rotation-resistant steel rope for winch and

hoisting arrangements. Where the choice of rope is concerned, DNV GL 2.22 specifies that

the manufacturer should be involved to ensure that a product with the right properties is used.

This was not done with the design for the HTV Eagle crane on GFB.

During its review of project documentation, including the FMECA, the team could not see

that the steel rope was identified or classified as a safety-critical component in the hoisting

arrangement. Nor were there any indications that service life calculations had been carried out

for components in the hoisting arrangement, including the rope.

4.2.2 Failure to follow up earlier incidents

On 1 January 2016, a birdcage was discovered on the steel rope for the HTV Eagle crane, and

broken strands were found on it. KCAD and MH Wirth concluded that the wrong rope type

had been installed, since right-hand lay rope was installed on a drum for left-hand lay. The

replacement was left-hand lay, but differed in tensile strength from the previous rope. No

assessment was made of whether this rope grade was correct for the crane arrangement, and

no particular inspection or maintenance of the rope was subsequently carried out.

The technical report subsequently published by MH Wirth does not reflect on other causes of

the incident than the rope having the wrong lay. A report from Kverneland Aqua AS to MH

Wirth states: “as installed, the rope will have S bends. Such bends could help to limit the

service life of the rope considerably – particularly with small rope-sheave dimensions.”

This did not result in any documented assessment of the rope’s service life. The report from

Kverneland Aqua AS also recommended a D/d ratio of 26, but this was again not taken into

account by MH Wirth nor communicated to Statoil or KCAD.

On 30 July 2016, the roller bearing in S1 on the HTV Eagle crane failed. Synergi contains the

following observation: “The winch is installed in such a way that the drum is at right-angles

to the rope sheave. In other words, the fleet angle of the steel rope across the winch drum

imposes transverse forces on the sheave/bearing. The preventive measure here is to replace

the sheave every six months.”

16

Neither KCAD nor Statoil assessed the possible consequences of this arrangement for the

rope.

These incidents ought to have initiated new follow-up routines and better methods for

inspection and maintenance of components in the hoisting arrangement. This would probably

also have identified that the design was deficient.

4.2.3 Insufficient knowledge of and expertise on steel rope

Statoil, Aibel, MH Wirth and KCAD all failed to assess the steel rope as a safety critical

component, even though it is solely responsible for supporting the weight of the boom and

load.

The boom weighs about 14.4 tonnes without load and can lift 3.5 tonnes. This means the rope

will be under a continuous load of at least 14 tonnes when the crane is not parked on the pipe

deck.

The rope used was a rotation-resistant type with high tensile strength, so that a rope with a

relatively small diameter can cope with a high load. The disadvantage is that high tensile

strength makes the rope stiff and unsuitable for a sheave diameter of 18xD/d.

In addition, checking a rotation-resistant steel rope for fatigue is very difficult. Fatigue caused

by large bending loads over small sheaves will cause strand fractures in the core which are not

detectable by visual checks. After the incident, KCAD incorporated visual checks of the steel

rope as a corrective measure in the maintenance system. However, this form of inspection has

its limits for detecting internal wear, broken strands and fatigue because they occur in the rope

core. This was demonstrated when the rope was taken out in January 2017 in order to install a

new sheave. Following a visual inspection, it was reinstalled despite having almost certainly

already reached the end of its fatigue life and being unacceptable for reinstallation.

4.2.4 Monitoring/inspection of steel rope

KCAD had only established visual checks of the rope, even though it is almost impossible to

detect fatigue in rotation-resistant steel rope in this way. Only external damage or wear would

be identified visually.

No system or routines were established to monitor the service life of or the number of load

cycles for the rope, even though the system is designed to be able to register the number of

lifting and lowering cycles for the boom. Only a replacement interval of 1.5 to two years is

described. Nor was provision made for assessing the type of rope (rotation-resistant) or its

sheave arrangement, or the pattern of use and operating time for the crane.

The PSA was told in interviews that the lists issued for short-term maintenance/lubrication

routines were so extensive that completing them within the allotted time was challenging. In

addition, the language in maintenance and inspection routines was too generalised to ensure

consistent implementation. The daily routine for the crane, for example, was described as

“Carry out a pre-use check of the lifting equipment”, and “Check the rope for damage and

deformation before use and during operation”. Nothing was said about particular wear sites or

vulnerable areas which called for special attention. Nor was provision made for access to the

most vulnerable areas.

17

Inspection and maintenance routines are vague and do not describe how the work is to be

done. As a result, inspection and maintenance of the steel rope can become dependent on the

person concerned and ambiguous. The work description for the rope in the maintenance

procedure for crane inspection, which is to be carried out every second, sixth and 12th month,

states: “inspect the rope for damage and wear in accordance with the applicable

requirements”. However, what requirements have been set by KCAD for inspecting the rope

are not specified in more detail.

No dimensional measurements of the rope were made on installation, nor when it was taken

off and inspected in January 2017. This would not have had an effect on the installed rope, but

maintenance requirements must be tailored to the equipment.

5 Potential of the incident

5.1 Actual consequence

The actual consequence of the crane boom falling onto the pipe deck was substantial material

damage to the crane and to the cable tray for the traverse crane on this deck. In addition, the

incident caused a halt to activity. Statoil has estimated the total financial cost at NOK 66

million.

Figure 10 The crane with bent crane arm and broken gripper arrangement.

Figure 10 shows the outer boom segment of the crane after it fell to the pipe deck. The

segment (striped red and yellow) has been bent by 90 degrees. The cable tray for the traverse

crane is deformed, while the gripper arrangement for loads hangs over the edge of the deck.

The railing broke off and fell towards the BOP deck, where it landed in the safety net. Figures

11–13 show other material damage.

Figure 11 The outer jib with the broken and detached gripper arrangement.

The gripper arrangement has been fractured.

One of the grippers is detached and lying on

the other side of the outer jig.

The crane with a bent outer boom segment.

The detached gripper arrangement

after it broke in two.

18

Figure 12 Photograph taken from below, under the pipe deck. Damage after the crane boom hit the deck.

Figure 13 The railing was broken off and fell towards the BOP deck, where it landed in the safety net.

5.2 Potential consequence

Under slightly different circumstances, the incident could have caused serious personal

injuries with the potential for a fatal outcome. A few minutes earlier, two people had

unhooked a load and adjusted the grippers on the gripper yoke for 3 ½-inch drill pipe. The

deck under the crane had been washed earlier the same day.

Had the derrick been positioned on a well slot further from the pipe deck, the boom could

have fallen straight down to the hatch/BOP deck and caused substantial material damage,

including to the well control equipment.

6 Observations

The PSA’s observations fall into two principal categories.

Nonconformities: observations where the PSA establishes a breach of/failure to comply

with the regulations.

Improvement points: observations where the PSA believes a breach of/failure to comply

with the regulations has occurred, but lacks sufficient information to establish this.

Damage to the underside of the

pipe deck after the crane boom fell.

Railing caught by the safety net over

the BOP deck.

19

6.1 Nonconformities

6.1.1 Risk assessment of equipment

Nonconformity Inadequate identification of risk.

Grounds The responsible party must choose technical, operational and organisational solutions which

reduce the probability that damage, faults, hazards or accidents will occur.

It emerged from the document review and conversations that neither Statoil nor MH Wirth

had assessed the steel rope as a safety-critical component, even though it alone bears the

whole weight of the crane boom and load. A single failure, such as the rope breaking, can

accordingly have big consequences.

Statoil defined the HTV Eagle cranes as a lifting facility with a low risk level, since it was

regarded as well-proven technology. Its certification is based on a low-level verification level

as defined in DNV-OSS-308, despite a number of design changes from the HTV Eagle crane

on Gullfaks C.

The FMECA did not identify and classify the steel rope as a safety-critical component in the

system, even though activities conducted on the drill floor and pipe deck meant that personnel

had to work beneath and close to the crane – when hooking loads on and off, for example, and

adjusting the gripper yoke. Nothing prevents the crane falling down if a single failure occurs

with load-carrying components during these activities. Nor is it possible to stop the crane

falling down when it is parked in the derrick.

Requirements

Section 4 of the management regulations on risk reduction

Section 17 of the management regulations on risk analyses and emergency preparedness

assessments

Section 5 of the facilities regulations on design of facilities

6.1.2 Investigation of and improvement measures following earlier incidents

Nonconformity Inadequate identification, investigation and follow-up of previous hazards and accidents.

Grounds The responsible party must ensure that hazards and accidents which could or have caused

acute pollution or other damage are registered and investigated to prevent repetition.

Earlier incidents with this crane where that had not been done emerged during the

investigation.

On 1 January 2016, a birdcage was discovered on the steel rope for the HTV Eagle crane, and

broken strands were found on it. KCAD and MH Wirth concluded that the wrong rope type

had been installed, since right-hand lay rope was installed on a drum for left-hand lay. The

replacement was left-hand lay, but differed in tensile strength from the previous rope. No

20

assessment was made of whether this rope grade was correct for the crane arrangement, and

no particular inspection or maintenance of the rope was subsequently carried out.

The technical report subsequently published by MH Wirth does not reflect on other causes of

the incident than the rope having the wrong lay. A report from Kverneland Aqua AS to MH

Wirth states: “as installed, the rope will have S bends. Such bends could help to limit the

service life of the rope considerably – particularly with small rope-sheave dimensions.”

This did not result in any documented assessment of the rope’s service life. The report from

Kverneland Aqua AS also recommended a D/d ratio of 26, but this was again not taken into

account by MH Wirth nor communicated to Statoil or KCAD.

On 30 July 2016, the roller bearing in S1 on the HTV Eagle crane failed. Synergi contains the

following observation: “The winch is installed in such a way that the drum is at right-angles

to the rope sheave. In other words, the fleet angle of the steel rope across the winch drum

imposes transverse forces on the sheave/bearing. The preventive measure here is to replace

the sheave every six months.”

Neither KCAD nor Statoil assessed the possible consequences of this arrangement for the

rope.

KCAD asked MH Wirth to verify that the sheave would cope with the oblique pull. This was

assessed and confirmed, but MH Wirth did not evaluate the possible consequences of the

arrangement for the steel rope.

Crane and lifting specialists in Statoil and at KCAD had expressed concern over the fleet

angle and wear on the sheave which could affect the rope, but these comments were not

followed up in a systematic manner.

In other words, the incidents with the rope and the failure of the roller bearing in the sheave

were not adequately investigated to prevent recurrence. Only minor changes were made to the

maintenance programme by expanding the visual inspection, but this cannot identify fatigue

in rotation-resistant steel rope

Requirements

Section 20, paragraph 1 of the management regulations on registration, review and

investigation of hazard and accident situations

Section 19, paragraph 1, litera e of the management regulations on collection, processing and

use of data

6.1.3 Utilisation of expertise

Nonconformity Insufficient use of crane and lifting expertise when specifying crane requirements.

Grounds Before decisions are taken, the responsible party must ensure that issues related to health,

safety and the environment have been comprehensively and adequately spelt out. That was

not the case here with regard to the HTV Eagle crane (Statoil’s specifications).

21

Users of the HTV Eagle crane on Gullfaks C were involved in order to benefit from their

experience. One result was the requirement that the rope should be protected by incorporating

it in the crane’s H beams in order to reduce the threat of damage when lifting equipment to

the drill floor with the deck crane. That led in turn to a reduction in sheave diameter from

26xD/d on Gullfaks C to 18xD/d on GFB, without any assessment of what effect these

changes would have on the rope’s service life. This should have been identified by involving

relevant crane and lifting expertise in the project.

Requirements

Section 11 of the management regulations on the basis for making decisions and decision

criteria

See section 8 of the management regulations on internal requirements

Section 14 of the management regulations on manning and competence

6.1.4 Responsibility for taking delivery and operation of equipment

Nonconformity Inadequate follow-up and checks when taking delivery of and operating machinery and

equipment.

Grounds The operator must ensure that everyone who does work for it complies with the requirements

specified in the health, safety and environmental legislation.

Taking delivery

When taking delivery and before starting to use the machine/crane, Statoil’s responsibilities

include seeing to it that the machinery regulations have been complied with, and that

necessary documentation and documents – such as spare part lists, maintenance routines and

instruction manuals – have been provided. One finding from the investigation team’s

document review is that the spare parts list was deficient and that the maintenance routines

were incomplete and unsuitable for the steel rope in question. The operation manual, for

example, had specified a rope replacement interval of 1.5 to two years, but this was not based

on service life calculations. No service life calculations were made for the rope.

Operation

Statoil has a responsibility for following up the drilling contractor’s work on the facility,

where its enterprise of competence in turn has a responsibility for following up the

maintenance and checking of offshore lifting equipment. A number of incidents with the HTV

Eagle crane were handled by KCAD through improvements implemented in collaboration

with MH Wirth. These incidents were treated by KCAD as remedial work for MH Wirth.

KCAD and Statoil’s rig representative failed to appreciate that the relevant incidents with the

crane should have prompted extraordinary checks and follow-up by the enterprise of

competence (see Statoil doc OM 209). According to information from interviews, Statoil’s

enterprise of competence was not involved in preparing or implementing the measures taken.

The investigation team takes the view that Statoil, including its enterprise of competence, has

failed to comply with its own requirements for taking delivery and for follow-up of control

and maintenance of crane and lifting equipment on GFB.

22

Requirements

Section 20 of the activities regulations on start-up and operation of facilities

Section 7 of the framework regulations on responsibilities pursuant to these regulations

See section 18 of the framework regulations on qualification and follow-up of other

participants

6.1.5 Maintenance

Nonconformity Deficiencies in the maintenance programme for the steel rope.

Grounds Failure modes which could present health, safety or environmental risk must be

systematically prevented with the aid of a maintenance programme.

It emerged from conversations with personnel on GFB and the documentation review that the

maintenance programme for steel rope does not specify the methodology and tolerance

requirements when inspecting the rope.

The inspection scope in KCAD’s maintenance programme is limited to a visual check of the

rope, and no physical method is specified.

When the sheave was disassembled and replaced, the whole rope was inspected and approved

on GFB. What the inspection comprised is not recorded in the maintenance programme.

MH Wirth’s operational description (Doc 880015-HA-022) proposes a measurement for the

work done by the steel rope in section 10.3.2 (“Winch rope monitoring. In order to plan

maintenance more easily, the control system will log the hours in mode and metres run. These

values are shown in the service display for the HTV Eagle”). However, KCAD did not make

this part of the maintenance routines.

Requirement

Section 47 of the activities regulations on maintenance programme

7 Other comments

Poor visibility from crane cabin It was observed during the investigation that the window pane in the HTV Eagle crane cabin

was broken, and that visibility from the cabin was thereby sharply reduced. The pane was of

toughened glass, but had not fallen to the deck because it was laminated.

KCAD explained that the pane was registered as broken on 25 October 2016 (Synergi

1488683), but that the damage had not been repaired in 133 days. The crane had been used

without special restrictions over a long period.

8 Barrier assessment

A brief assessment has been made of the barriers which functioned and failed to function.

23

The barriers are assessed in relation to the technical, organisational and operational barrier

elements.

Non-functioning

barriers

Functioning

barriers

Technical

elements

Organisational

elements

Operational

elements

Technical Individual

faults can lead

to failure and

fault modes on

the HTV Eagle

crane.

No system

exists for

parking and

disengaging

the steel rope

when the crane

is parked in the

derrick.

Inadequate

knowledge of

steel rope.

Inaccurate

description of

maintenance

intervals in the

user manual.

Maintenance

system

Unclear

description of

inspection

routines and

maintenance

tasks.

Maintenance Inadequate

knowledge of

maintenance

requirements for

steel rope.

Follow-up/

learning from

earlier incidents

Responsibility for

the equipment

and for following

up incidents was

spread between

several companies

and people, and

became less clear.

Procedure for

cordoning off

the area under

the HTV

Eagle crane.

In accordance

with operational

routines.

9 Discussion of uncertainties

We have not identified any uncertainties with regard to the course of events or causes.

24

10 Statoil’s investigation report

The incident has been investigated by Statoil’s internal audit function (COA INV), and the

PSA received its report on 8 June.

This report’s description of the course of events and causes by and large coincides with the

investigation team’s findings and assessments.

A number of the recommended measures in the report are rather vague. Measures are phrased

in such a way as “secure that ...”, for example, without describing how this should be done or

what activities are to be performed. Nothing is said about a systematic review of similar crane

and lifting equipment with rotation-resistant rope.

Statoil has sent the damaged steel rope and sheave to W Giertsen Service AS for assessment

and investigation. The investigation team has taken note of the assessments of the actual and

potential consequences and the material technology investigation and analysis

11 Appendix

A: List of documents taken into account in the investigation.

B: Overview of personnel interviewed.